{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.933333333333334,
  "eval_steps": 150,
  "global_step": 335,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.014814814814814815,
      "grad_norm": 24.245065689086914,
      "learning_rate": 0.0,
      "loss": 1.5986,
      "step": 1
    },
    {
      "epoch": 0.02962962962962963,
      "grad_norm": 24.59937286376953,
      "learning_rate": 1.965616322328226e-06,
      "loss": 1.5387,
      "step": 2
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 22.017257690429688,
      "learning_rate": 3.1154281616956676e-06,
      "loss": 1.4677,
      "step": 3
    },
    {
      "epoch": 0.05925925925925926,
      "grad_norm": 19.979198455810547,
      "learning_rate": 3.931232644656452e-06,
      "loss": 1.3971,
      "step": 4
    },
    {
      "epoch": 0.07407407407407407,
      "grad_norm": 10.355873107910156,
      "learning_rate": 4.5640197625830816e-06,
      "loss": 1.1965,
      "step": 5
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 7.20529842376709,
      "learning_rate": 5.081044484023894e-06,
      "loss": 1.0327,
      "step": 6
    },
    {
      "epoch": 0.1037037037037037,
      "grad_norm": 6.543303489685059,
      "learning_rate": 5.518182657364912e-06,
      "loss": 1.0091,
      "step": 7
    },
    {
      "epoch": 0.11851851851851852,
      "grad_norm": 5.638514041900635,
      "learning_rate": 5.896848966984678e-06,
      "loss": 0.8685,
      "step": 8
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 4.796297073364258,
      "learning_rate": 6.230856323391335e-06,
      "loss": 0.816,
      "step": 9
    },
    {
      "epoch": 0.14814814814814814,
      "grad_norm": 6.3762431144714355,
      "learning_rate": 6.5296360849113085e-06,
      "loss": 0.8679,
      "step": 10
    },
    {
      "epoch": 0.16296296296296298,
      "grad_norm": 4.701241493225098,
      "learning_rate": 6.7999152555718276e-06,
      "loss": 0.7973,
      "step": 11
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 4.334161281585693,
      "learning_rate": 7.04666080635212e-06,
      "loss": 0.6289,
      "step": 12
    },
    {
      "epoch": 0.1925925925925926,
      "grad_norm": 2.9507179260253906,
      "learning_rate": 7.273644709769792e-06,
      "loss": 0.7202,
      "step": 13
    },
    {
      "epoch": 0.2074074074074074,
      "grad_norm": 3.3083744049072266,
      "learning_rate": 7.483798979693137e-06,
      "loss": 0.6864,
      "step": 14
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 3.1791484355926514,
      "learning_rate": 7.679447924278749e-06,
      "loss": 0.7458,
      "step": 15
    },
    {
      "epoch": 0.23703703703703705,
      "grad_norm": 2.6846699714660645,
      "learning_rate": 7.862465289312904e-06,
      "loss": 0.6543,
      "step": 16
    },
    {
      "epoch": 0.2518518518518518,
      "grad_norm": 3.1359522342681885,
      "learning_rate": 8.034383677671775e-06,
      "loss": 0.531,
      "step": 17
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 3.0577006340026855,
      "learning_rate": 8.19647264571956e-06,
      "loss": 0.6345,
      "step": 18
    },
    {
      "epoch": 0.2814814814814815,
      "grad_norm": 3.066653251647949,
      "learning_rate": 8.349795656491867e-06,
      "loss": 0.5942,
      "step": 19
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 3.3706552982330322,
      "learning_rate": 8.495252407239533e-06,
      "loss": 0.638,
      "step": 20
    },
    {
      "epoch": 0.3111111111111111,
      "grad_norm": 2.55362606048584,
      "learning_rate": 8.633610819060579e-06,
      "loss": 0.602,
      "step": 21
    },
    {
      "epoch": 0.32592592592592595,
      "grad_norm": 2.864042282104492,
      "learning_rate": 8.765531577900054e-06,
      "loss": 0.6849,
      "step": 22
    },
    {
      "epoch": 0.34074074074074073,
      "grad_norm": 2.357649326324463,
      "learning_rate": 8.891587215888663e-06,
      "loss": 0.5194,
      "step": 23
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 2.659815549850464,
      "learning_rate": 9.012277128680346e-06,
      "loss": 0.4647,
      "step": 24
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 3.5252552032470703,
      "learning_rate": 9.128039525166163e-06,
      "loss": 0.6621,
      "step": 25
    },
    {
      "epoch": 0.3851851851851852,
      "grad_norm": 2.833387613296509,
      "learning_rate": 9.239261032098019e-06,
      "loss": 0.5758,
      "step": 26
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.5673093795776367,
      "learning_rate": 9.346284485087002e-06,
      "loss": 0.4742,
      "step": 27
    },
    {
      "epoch": 0.4148148148148148,
      "grad_norm": 3.8754830360412598,
      "learning_rate": 9.449415302021363e-06,
      "loss": 0.5632,
      "step": 28
    },
    {
      "epoch": 0.42962962962962964,
      "grad_norm": 2.8832461833953857,
      "learning_rate": 9.548926737583076e-06,
      "loss": 0.5039,
      "step": 29
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 2.340471029281616,
      "learning_rate": 9.645064246606975e-06,
      "loss": 0.4832,
      "step": 30
    },
    {
      "epoch": 0.45925925925925926,
      "grad_norm": 3.4840638637542725,
      "learning_rate": 9.738049131714717e-06,
      "loss": 0.6216,
      "step": 31
    },
    {
      "epoch": 0.4740740740740741,
      "grad_norm": 1.927802562713623,
      "learning_rate": 9.828081611641132e-06,
      "loss": 0.4316,
      "step": 32
    },
    {
      "epoch": 0.4888888888888889,
      "grad_norm": 1.9146533012390137,
      "learning_rate": 9.915343417267494e-06,
      "loss": 0.4261,
      "step": 33
    },
    {
      "epoch": 0.5037037037037037,
      "grad_norm": 2.636239767074585,
      "learning_rate": 1e-05,
      "loss": 0.4445,
      "step": 34
    },
    {
      "epoch": 0.5185185185185185,
      "grad_norm": 2.7775421142578125,
      "learning_rate": 1e-05,
      "loss": 0.5895,
      "step": 35
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 3.3831987380981445,
      "learning_rate": 9.966777408637874e-06,
      "loss": 0.5247,
      "step": 36
    },
    {
      "epoch": 0.5481481481481482,
      "grad_norm": 2.883329391479492,
      "learning_rate": 9.933554817275748e-06,
      "loss": 0.4712,
      "step": 37
    },
    {
      "epoch": 0.562962962962963,
      "grad_norm": 2.908726215362549,
      "learning_rate": 9.900332225913623e-06,
      "loss": 0.4991,
      "step": 38
    },
    {
      "epoch": 0.5777777777777777,
      "grad_norm": 2.307657480239868,
      "learning_rate": 9.867109634551495e-06,
      "loss": 0.4174,
      "step": 39
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 2.3641955852508545,
      "learning_rate": 9.83388704318937e-06,
      "loss": 0.5014,
      "step": 40
    },
    {
      "epoch": 0.6074074074074074,
      "grad_norm": 2.850147247314453,
      "learning_rate": 9.800664451827243e-06,
      "loss": 0.5025,
      "step": 41
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 2.9469687938690186,
      "learning_rate": 9.767441860465117e-06,
      "loss": 0.4127,
      "step": 42
    },
    {
      "epoch": 0.6370370370370371,
      "grad_norm": 2.0221896171569824,
      "learning_rate": 9.734219269102992e-06,
      "loss": 0.4793,
      "step": 43
    },
    {
      "epoch": 0.6518518518518519,
      "grad_norm": 2.2807424068450928,
      "learning_rate": 9.700996677740865e-06,
      "loss": 0.4363,
      "step": 44
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.9812514781951904,
      "learning_rate": 9.66777408637874e-06,
      "loss": 0.472,
      "step": 45
    },
    {
      "epoch": 0.6814814814814815,
      "grad_norm": 2.3653407096862793,
      "learning_rate": 9.634551495016612e-06,
      "loss": 0.4747,
      "step": 46
    },
    {
      "epoch": 0.6962962962962963,
      "grad_norm": 1.875173807144165,
      "learning_rate": 9.601328903654485e-06,
      "loss": 0.4466,
      "step": 47
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 2.5644772052764893,
      "learning_rate": 9.56810631229236e-06,
      "loss": 0.4908,
      "step": 48
    },
    {
      "epoch": 0.725925925925926,
      "grad_norm": 2.5701663494110107,
      "learning_rate": 9.534883720930234e-06,
      "loss": 0.5709,
      "step": 49
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 3.4693243503570557,
      "learning_rate": 9.501661129568107e-06,
      "loss": 0.4436,
      "step": 50
    },
    {
      "epoch": 0.7555555555555555,
      "grad_norm": 1.7804033756256104,
      "learning_rate": 9.468438538205981e-06,
      "loss": 0.5096,
      "step": 51
    },
    {
      "epoch": 0.7703703703703704,
      "grad_norm": 2.171896457672119,
      "learning_rate": 9.435215946843854e-06,
      "loss": 0.4063,
      "step": 52
    },
    {
      "epoch": 0.7851851851851852,
      "grad_norm": 2.830214500427246,
      "learning_rate": 9.401993355481728e-06,
      "loss": 0.6981,
      "step": 53
    },
    {
      "epoch": 0.8,
      "grad_norm": 3.3130273818969727,
      "learning_rate": 9.368770764119603e-06,
      "loss": 0.5553,
      "step": 54
    },
    {
      "epoch": 0.8148148148148148,
      "grad_norm": 2.5522916316986084,
      "learning_rate": 9.335548172757476e-06,
      "loss": 0.5471,
      "step": 55
    },
    {
      "epoch": 0.8296296296296296,
      "grad_norm": 1.8053007125854492,
      "learning_rate": 9.30232558139535e-06,
      "loss": 0.4449,
      "step": 56
    },
    {
      "epoch": 0.8444444444444444,
      "grad_norm": 2.2297627925872803,
      "learning_rate": 9.269102990033223e-06,
      "loss": 0.4782,
      "step": 57
    },
    {
      "epoch": 0.8592592592592593,
      "grad_norm": 2.388592004776001,
      "learning_rate": 9.235880398671098e-06,
      "loss": 0.4333,
      "step": 58
    },
    {
      "epoch": 0.8740740740740741,
      "grad_norm": 2.2160580158233643,
      "learning_rate": 9.20265780730897e-06,
      "loss": 0.397,
      "step": 59
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 2.1153295040130615,
      "learning_rate": 9.169435215946845e-06,
      "loss": 0.4976,
      "step": 60
    },
    {
      "epoch": 0.9037037037037037,
      "grad_norm": 2.1571428775787354,
      "learning_rate": 9.136212624584718e-06,
      "loss": 0.5009,
      "step": 61
    },
    {
      "epoch": 0.9185185185185185,
      "grad_norm": 1.9368093013763428,
      "learning_rate": 9.102990033222592e-06,
      "loss": 0.4225,
      "step": 62
    },
    {
      "epoch": 0.9333333333333333,
      "grad_norm": 1.793280839920044,
      "learning_rate": 9.069767441860465e-06,
      "loss": 0.4407,
      "step": 63
    },
    {
      "epoch": 0.9481481481481482,
      "grad_norm": 1.7864512205123901,
      "learning_rate": 9.03654485049834e-06,
      "loss": 0.481,
      "step": 64
    },
    {
      "epoch": 0.9629629629629629,
      "grad_norm": 2.982163429260254,
      "learning_rate": 9.003322259136214e-06,
      "loss": 0.5735,
      "step": 65
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 2.1836278438568115,
      "learning_rate": 8.970099667774087e-06,
      "loss": 0.4532,
      "step": 66
    },
    {
      "epoch": 0.9925925925925926,
      "grad_norm": 2.726743221282959,
      "learning_rate": 8.93687707641196e-06,
      "loss": 0.5093,
      "step": 67
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.726743221282959,
      "learning_rate": 8.93687707641196e-06,
      "loss": 0.4299,
      "step": 68
    },
    {
      "epoch": 1.0148148148148148,
      "grad_norm": 2.7928483486175537,
      "learning_rate": 8.903654485049834e-06,
      "loss": 0.3207,
      "step": 69
    },
    {
      "epoch": 1.0296296296296297,
      "grad_norm": 1.4182758331298828,
      "learning_rate": 8.870431893687709e-06,
      "loss": 0.4154,
      "step": 70
    },
    {
      "epoch": 1.0444444444444445,
      "grad_norm": 2.5209672451019287,
      "learning_rate": 8.837209302325582e-06,
      "loss": 0.4017,
      "step": 71
    },
    {
      "epoch": 1.0592592592592593,
      "grad_norm": 1.528294324874878,
      "learning_rate": 8.803986710963456e-06,
      "loss": 0.3515,
      "step": 72
    },
    {
      "epoch": 1.074074074074074,
      "grad_norm": 2.7746169567108154,
      "learning_rate": 8.770764119601329e-06,
      "loss": 0.4338,
      "step": 73
    },
    {
      "epoch": 1.0888888888888888,
      "grad_norm": 1.9682352542877197,
      "learning_rate": 8.737541528239203e-06,
      "loss": 0.4154,
      "step": 74
    },
    {
      "epoch": 1.1037037037037036,
      "grad_norm": 2.2559337615966797,
      "learning_rate": 8.704318936877078e-06,
      "loss": 0.368,
      "step": 75
    },
    {
      "epoch": 1.1185185185185185,
      "grad_norm": 1.3220306634902954,
      "learning_rate": 8.67109634551495e-06,
      "loss": 0.3142,
      "step": 76
    },
    {
      "epoch": 1.1333333333333333,
      "grad_norm": 2.455972671508789,
      "learning_rate": 8.637873754152825e-06,
      "loss": 0.4777,
      "step": 77
    },
    {
      "epoch": 1.1481481481481481,
      "grad_norm": 2.32432222366333,
      "learning_rate": 8.604651162790698e-06,
      "loss": 0.4585,
      "step": 78
    },
    {
      "epoch": 1.162962962962963,
      "grad_norm": 2.2095751762390137,
      "learning_rate": 8.571428571428571e-06,
      "loss": 0.4513,
      "step": 79
    },
    {
      "epoch": 1.1777777777777778,
      "grad_norm": 2.29775071144104,
      "learning_rate": 8.538205980066447e-06,
      "loss": 0.3315,
      "step": 80
    },
    {
      "epoch": 1.1925925925925926,
      "grad_norm": 2.390573740005493,
      "learning_rate": 8.50498338870432e-06,
      "loss": 0.4328,
      "step": 81
    },
    {
      "epoch": 1.2074074074074075,
      "grad_norm": 2.1682214736938477,
      "learning_rate": 8.471760797342193e-06,
      "loss": 0.3708,
      "step": 82
    },
    {
      "epoch": 1.2222222222222223,
      "grad_norm": 2.305809736251831,
      "learning_rate": 8.438538205980067e-06,
      "loss": 0.333,
      "step": 83
    },
    {
      "epoch": 1.237037037037037,
      "grad_norm": 1.9154317378997803,
      "learning_rate": 8.40531561461794e-06,
      "loss": 0.3514,
      "step": 84
    },
    {
      "epoch": 1.2518518518518518,
      "grad_norm": 2.1949751377105713,
      "learning_rate": 8.372093023255815e-06,
      "loss": 0.3423,
      "step": 85
    },
    {
      "epoch": 1.2666666666666666,
      "grad_norm": 1.4846103191375732,
      "learning_rate": 8.338870431893689e-06,
      "loss": 0.3935,
      "step": 86
    },
    {
      "epoch": 1.2814814814814814,
      "grad_norm": 2.3148951530456543,
      "learning_rate": 8.305647840531562e-06,
      "loss": 0.3683,
      "step": 87
    },
    {
      "epoch": 1.2962962962962963,
      "grad_norm": 1.7715332508087158,
      "learning_rate": 8.272425249169436e-06,
      "loss": 0.3415,
      "step": 88
    },
    {
      "epoch": 1.3111111111111111,
      "grad_norm": 1.6804319620132446,
      "learning_rate": 8.23920265780731e-06,
      "loss": 0.4075,
      "step": 89
    },
    {
      "epoch": 1.325925925925926,
      "grad_norm": 1.976668119430542,
      "learning_rate": 8.205980066445184e-06,
      "loss": 0.295,
      "step": 90
    },
    {
      "epoch": 1.3407407407407408,
      "grad_norm": 1.8023229837417603,
      "learning_rate": 8.172757475083057e-06,
      "loss": 0.389,
      "step": 91
    },
    {
      "epoch": 1.3555555555555556,
      "grad_norm": 1.7759190797805786,
      "learning_rate": 8.139534883720931e-06,
      "loss": 0.3517,
      "step": 92
    },
    {
      "epoch": 1.3703703703703702,
      "grad_norm": 1.3482650518417358,
      "learning_rate": 8.106312292358804e-06,
      "loss": 0.3606,
      "step": 93
    },
    {
      "epoch": 1.3851851851851853,
      "grad_norm": 2.5421998500823975,
      "learning_rate": 8.073089700996678e-06,
      "loss": 0.5241,
      "step": 94
    },
    {
      "epoch": 1.4,
      "grad_norm": 3.181865692138672,
      "learning_rate": 8.039867109634553e-06,
      "loss": 0.375,
      "step": 95
    },
    {
      "epoch": 1.4148148148148147,
      "grad_norm": 1.7613658905029297,
      "learning_rate": 8.006644518272426e-06,
      "loss": 0.4339,
      "step": 96
    },
    {
      "epoch": 1.4296296296296296,
      "grad_norm": 1.4631813764572144,
      "learning_rate": 7.9734219269103e-06,
      "loss": 0.3762,
      "step": 97
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 1.71363365650177,
      "learning_rate": 7.940199335548173e-06,
      "loss": 0.3224,
      "step": 98
    },
    {
      "epoch": 1.4592592592592593,
      "grad_norm": 2.1795740127563477,
      "learning_rate": 7.906976744186048e-06,
      "loss": 0.3816,
      "step": 99
    },
    {
      "epoch": 1.474074074074074,
      "grad_norm": 1.5822690725326538,
      "learning_rate": 7.873754152823922e-06,
      "loss": 0.4523,
      "step": 100
    },
    {
      "epoch": 1.488888888888889,
      "grad_norm": 1.7851877212524414,
      "learning_rate": 7.840531561461795e-06,
      "loss": 0.3665,
      "step": 101
    },
    {
      "epoch": 1.5037037037037035,
      "grad_norm": 1.671633243560791,
      "learning_rate": 7.807308970099668e-06,
      "loss": 0.3217,
      "step": 102
    },
    {
      "epoch": 1.5185185185185186,
      "grad_norm": 2.666489601135254,
      "learning_rate": 7.774086378737542e-06,
      "loss": 0.4801,
      "step": 103
    },
    {
      "epoch": 1.5333333333333332,
      "grad_norm": 1.3500992059707642,
      "learning_rate": 7.740863787375415e-06,
      "loss": 0.3407,
      "step": 104
    },
    {
      "epoch": 1.5481481481481483,
      "grad_norm": 1.4269295930862427,
      "learning_rate": 7.70764119601329e-06,
      "loss": 0.3082,
      "step": 105
    },
    {
      "epoch": 1.5629629629629629,
      "grad_norm": 1.4882391691207886,
      "learning_rate": 7.674418604651164e-06,
      "loss": 0.3497,
      "step": 106
    },
    {
      "epoch": 1.5777777777777777,
      "grad_norm": 1.4180535078048706,
      "learning_rate": 7.641196013289037e-06,
      "loss": 0.3963,
      "step": 107
    },
    {
      "epoch": 1.5925925925925926,
      "grad_norm": 1.4838252067565918,
      "learning_rate": 7.6079734219269106e-06,
      "loss": 0.3956,
      "step": 108
    },
    {
      "epoch": 1.6074074074074074,
      "grad_norm": 2.0815930366516113,
      "learning_rate": 7.574750830564784e-06,
      "loss": 0.4667,
      "step": 109
    },
    {
      "epoch": 1.6222222222222222,
      "grad_norm": 1.80450439453125,
      "learning_rate": 7.541528239202659e-06,
      "loss": 0.5329,
      "step": 110
    },
    {
      "epoch": 1.637037037037037,
      "grad_norm": 2.5666868686676025,
      "learning_rate": 7.508305647840532e-06,
      "loss": 0.5433,
      "step": 111
    },
    {
      "epoch": 1.651851851851852,
      "grad_norm": 1.733236312866211,
      "learning_rate": 7.475083056478406e-06,
      "loss": 0.4304,
      "step": 112
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 2.059544086456299,
      "learning_rate": 7.44186046511628e-06,
      "loss": 0.3601,
      "step": 113
    },
    {
      "epoch": 1.6814814814814816,
      "grad_norm": 1.8904452323913574,
      "learning_rate": 7.408637873754153e-06,
      "loss": 0.385,
      "step": 114
    },
    {
      "epoch": 1.6962962962962962,
      "grad_norm": 1.3712379932403564,
      "learning_rate": 7.375415282392027e-06,
      "loss": 0.2972,
      "step": 115
    },
    {
      "epoch": 1.7111111111111112,
      "grad_norm": 1.5603693723678589,
      "learning_rate": 7.342192691029902e-06,
      "loss": 0.3953,
      "step": 116
    },
    {
      "epoch": 1.7259259259259259,
      "grad_norm": 1.4126758575439453,
      "learning_rate": 7.308970099667775e-06,
      "loss": 0.3582,
      "step": 117
    },
    {
      "epoch": 1.7407407407407407,
      "grad_norm": 2.0345335006713867,
      "learning_rate": 7.275747508305648e-06,
      "loss": 0.5015,
      "step": 118
    },
    {
      "epoch": 1.7555555555555555,
      "grad_norm": 2.6806693077087402,
      "learning_rate": 7.242524916943522e-06,
      "loss": 0.4006,
      "step": 119
    },
    {
      "epoch": 1.7703703703703704,
      "grad_norm": 1.6245332956314087,
      "learning_rate": 7.209302325581395e-06,
      "loss": 0.4188,
      "step": 120
    },
    {
      "epoch": 1.7851851851851852,
      "grad_norm": 1.6593953371047974,
      "learning_rate": 7.17607973421927e-06,
      "loss": 0.3694,
      "step": 121
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.595201849937439,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 0.2838,
      "step": 122
    },
    {
      "epoch": 1.8148148148148149,
      "grad_norm": 1.921108603477478,
      "learning_rate": 7.109634551495017e-06,
      "loss": 0.3044,
      "step": 123
    },
    {
      "epoch": 1.8296296296296295,
      "grad_norm": 1.5152816772460938,
      "learning_rate": 7.076411960132891e-06,
      "loss": 0.3705,
      "step": 124
    },
    {
      "epoch": 1.8444444444444446,
      "grad_norm": 1.7763299942016602,
      "learning_rate": 7.0431893687707646e-06,
      "loss": 0.3935,
      "step": 125
    },
    {
      "epoch": 1.8592592592592592,
      "grad_norm": 1.3404872417449951,
      "learning_rate": 7.009966777408639e-06,
      "loss": 0.3951,
      "step": 126
    },
    {
      "epoch": 1.8740740740740742,
      "grad_norm": 1.7480947971343994,
      "learning_rate": 6.976744186046513e-06,
      "loss": 0.3694,
      "step": 127
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 1.3903143405914307,
      "learning_rate": 6.9435215946843855e-06,
      "loss": 0.3963,
      "step": 128
    },
    {
      "epoch": 1.9037037037037037,
      "grad_norm": 1.579624056816101,
      "learning_rate": 6.910299003322259e-06,
      "loss": 0.4147,
      "step": 129
    },
    {
      "epoch": 1.9185185185185185,
      "grad_norm": 1.6633076667785645,
      "learning_rate": 6.877076411960133e-06,
      "loss": 0.415,
      "step": 130
    },
    {
      "epoch": 1.9333333333333333,
      "grad_norm": 1.4897836446762085,
      "learning_rate": 6.843853820598007e-06,
      "loss": 0.3438,
      "step": 131
    },
    {
      "epoch": 1.9481481481481482,
      "grad_norm": 1.7405959367752075,
      "learning_rate": 6.810631229235881e-06,
      "loss": 0.3973,
      "step": 132
    },
    {
      "epoch": 1.9629629629629628,
      "grad_norm": 1.6084774732589722,
      "learning_rate": 6.777408637873755e-06,
      "loss": 0.3904,
      "step": 133
    },
    {
      "epoch": 1.9777777777777779,
      "grad_norm": 1.4988147020339966,
      "learning_rate": 6.744186046511628e-06,
      "loss": 0.4101,
      "step": 134
    },
    {
      "epoch": 1.9925925925925925,
      "grad_norm": 2.449335813522339,
      "learning_rate": 6.710963455149502e-06,
      "loss": 0.3886,
      "step": 135
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.2566375732421875,
      "learning_rate": 6.6777408637873766e-06,
      "loss": 0.3651,
      "step": 136
    },
    {
      "epoch": 2.0148148148148146,
      "grad_norm": 1.5977928638458252,
      "learning_rate": 6.64451827242525e-06,
      "loss": 0.3709,
      "step": 137
    },
    {
      "epoch": 2.0296296296296297,
      "grad_norm": 2.1500344276428223,
      "learning_rate": 6.611295681063124e-06,
      "loss": 0.4124,
      "step": 138
    },
    {
      "epoch": 2.0444444444444443,
      "grad_norm": 1.3736720085144043,
      "learning_rate": 6.578073089700997e-06,
      "loss": 0.3334,
      "step": 139
    },
    {
      "epoch": 2.0592592592592593,
      "grad_norm": 1.6602332592010498,
      "learning_rate": 6.54485049833887e-06,
      "loss": 0.3443,
      "step": 140
    },
    {
      "epoch": 2.074074074074074,
      "grad_norm": 1.5972793102264404,
      "learning_rate": 6.511627906976745e-06,
      "loss": 0.3969,
      "step": 141
    },
    {
      "epoch": 2.088888888888889,
      "grad_norm": 1.5588147640228271,
      "learning_rate": 6.4784053156146185e-06,
      "loss": 0.2632,
      "step": 142
    },
    {
      "epoch": 2.1037037037037036,
      "grad_norm": 1.4944829940795898,
      "learning_rate": 6.445182724252492e-06,
      "loss": 0.3403,
      "step": 143
    },
    {
      "epoch": 2.1185185185185187,
      "grad_norm": 1.799228310585022,
      "learning_rate": 6.411960132890366e-06,
      "loss": 0.3669,
      "step": 144
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 1.5002415180206299,
      "learning_rate": 6.3787375415282395e-06,
      "loss": 0.3037,
      "step": 145
    },
    {
      "epoch": 2.148148148148148,
      "grad_norm": 1.5463218688964844,
      "learning_rate": 6.345514950166114e-06,
      "loss": 0.2999,
      "step": 146
    },
    {
      "epoch": 2.162962962962963,
      "grad_norm": 1.8339827060699463,
      "learning_rate": 6.312292358803988e-06,
      "loss": 0.3781,
      "step": 147
    },
    {
      "epoch": 2.1777777777777776,
      "grad_norm": 1.9693403244018555,
      "learning_rate": 6.279069767441861e-06,
      "loss": 0.3527,
      "step": 148
    },
    {
      "epoch": 2.1925925925925926,
      "grad_norm": 1.3716813325881958,
      "learning_rate": 6.245847176079734e-06,
      "loss": 0.3193,
      "step": 149
    },
    {
      "epoch": 2.2074074074074073,
      "grad_norm": 1.2879470586776733,
      "learning_rate": 6.212624584717608e-06,
      "loss": 0.3079,
      "step": 150
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 1.3284484148025513,
      "learning_rate": 6.179401993355482e-06,
      "loss": 0.3552,
      "step": 151
    },
    {
      "epoch": 2.237037037037037,
      "grad_norm": 1.1561126708984375,
      "learning_rate": 6.146179401993356e-06,
      "loss": 0.2879,
      "step": 152
    },
    {
      "epoch": 2.251851851851852,
      "grad_norm": 1.2936991453170776,
      "learning_rate": 6.11295681063123e-06,
      "loss": 0.2926,
      "step": 153
    },
    {
      "epoch": 2.2666666666666666,
      "grad_norm": 1.5603609085083008,
      "learning_rate": 6.079734219269103e-06,
      "loss": 0.3591,
      "step": 154
    },
    {
      "epoch": 2.2814814814814817,
      "grad_norm": 1.408933162689209,
      "learning_rate": 6.046511627906977e-06,
      "loss": 0.3313,
      "step": 155
    },
    {
      "epoch": 2.2962962962962963,
      "grad_norm": 1.9709101915359497,
      "learning_rate": 6.0132890365448515e-06,
      "loss": 0.3948,
      "step": 156
    },
    {
      "epoch": 2.311111111111111,
      "grad_norm": 1.5349321365356445,
      "learning_rate": 5.980066445182725e-06,
      "loss": 0.3227,
      "step": 157
    },
    {
      "epoch": 2.325925925925926,
      "grad_norm": 1.5960361957550049,
      "learning_rate": 5.946843853820599e-06,
      "loss": 0.3532,
      "step": 158
    },
    {
      "epoch": 2.3407407407407406,
      "grad_norm": 1.8431493043899536,
      "learning_rate": 5.9136212624584725e-06,
      "loss": 0.3621,
      "step": 159
    },
    {
      "epoch": 2.3555555555555556,
      "grad_norm": 1.875801920890808,
      "learning_rate": 5.880398671096345e-06,
      "loss": 0.4026,
      "step": 160
    },
    {
      "epoch": 2.3703703703703702,
      "grad_norm": 1.7672104835510254,
      "learning_rate": 5.847176079734221e-06,
      "loss": 0.3889,
      "step": 161
    },
    {
      "epoch": 2.3851851851851853,
      "grad_norm": 1.5263267755508423,
      "learning_rate": 5.8139534883720935e-06,
      "loss": 0.332,
      "step": 162
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.624315857887268,
      "learning_rate": 5.780730897009967e-06,
      "loss": 0.3167,
      "step": 163
    },
    {
      "epoch": 2.414814814814815,
      "grad_norm": 1.9534611701965332,
      "learning_rate": 5.747508305647841e-06,
      "loss": 0.3733,
      "step": 164
    },
    {
      "epoch": 2.4296296296296296,
      "grad_norm": 1.6925545930862427,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.3631,
      "step": 165
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 1.6101276874542236,
      "learning_rate": 5.681063122923588e-06,
      "loss": 0.3561,
      "step": 166
    },
    {
      "epoch": 2.4592592592592593,
      "grad_norm": 1.5481266975402832,
      "learning_rate": 5.647840531561463e-06,
      "loss": 0.349,
      "step": 167
    },
    {
      "epoch": 2.474074074074074,
      "grad_norm": 1.6006097793579102,
      "learning_rate": 5.614617940199336e-06,
      "loss": 0.3862,
      "step": 168
    },
    {
      "epoch": 2.488888888888889,
      "grad_norm": 1.4279389381408691,
      "learning_rate": 5.58139534883721e-06,
      "loss": 0.3236,
      "step": 169
    },
    {
      "epoch": 2.5037037037037035,
      "grad_norm": 1.2405160665512085,
      "learning_rate": 5.548172757475083e-06,
      "loss": 0.3013,
      "step": 170
    },
    {
      "epoch": 2.5185185185185186,
      "grad_norm": 1.3768985271453857,
      "learning_rate": 5.5149501661129565e-06,
      "loss": 0.3626,
      "step": 171
    },
    {
      "epoch": 2.533333333333333,
      "grad_norm": 1.6408584117889404,
      "learning_rate": 5.481727574750831e-06,
      "loss": 0.3196,
      "step": 172
    },
    {
      "epoch": 2.5481481481481483,
      "grad_norm": 1.6769694089889526,
      "learning_rate": 5.448504983388705e-06,
      "loss": 0.3694,
      "step": 173
    },
    {
      "epoch": 2.562962962962963,
      "grad_norm": 1.3804996013641357,
      "learning_rate": 5.415282392026578e-06,
      "loss": 0.3502,
      "step": 174
    },
    {
      "epoch": 2.5777777777777775,
      "grad_norm": 1.4021036624908447,
      "learning_rate": 5.382059800664452e-06,
      "loss": 0.3306,
      "step": 175
    },
    {
      "epoch": 2.5925925925925926,
      "grad_norm": 1.2684727907180786,
      "learning_rate": 5.348837209302326e-06,
      "loss": 0.3106,
      "step": 176
    },
    {
      "epoch": 2.6074074074074076,
      "grad_norm": 1.5925308465957642,
      "learning_rate": 5.3156146179402e-06,
      "loss": 0.4211,
      "step": 177
    },
    {
      "epoch": 2.6222222222222222,
      "grad_norm": 1.5546250343322754,
      "learning_rate": 5.282392026578074e-06,
      "loss": 0.3301,
      "step": 178
    },
    {
      "epoch": 2.637037037037037,
      "grad_norm": 1.4001022577285767,
      "learning_rate": 5.2491694352159475e-06,
      "loss": 0.3746,
      "step": 179
    },
    {
      "epoch": 2.651851851851852,
      "grad_norm": 1.4525564908981323,
      "learning_rate": 5.215946843853821e-06,
      "loss": 0.2681,
      "step": 180
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 1.27578604221344,
      "learning_rate": 5.182724252491694e-06,
      "loss": 0.3106,
      "step": 181
    },
    {
      "epoch": 2.6814814814814816,
      "grad_norm": 1.4694350957870483,
      "learning_rate": 5.149501661129569e-06,
      "loss": 0.3065,
      "step": 182
    },
    {
      "epoch": 2.696296296296296,
      "grad_norm": 1.3258838653564453,
      "learning_rate": 5.116279069767442e-06,
      "loss": 0.3233,
      "step": 183
    },
    {
      "epoch": 2.7111111111111112,
      "grad_norm": 1.1991218328475952,
      "learning_rate": 5.083056478405316e-06,
      "loss": 0.2786,
      "step": 184
    },
    {
      "epoch": 2.725925925925926,
      "grad_norm": 1.134046196937561,
      "learning_rate": 5.0498338870431895e-06,
      "loss": 0.2788,
      "step": 185
    },
    {
      "epoch": 2.7407407407407405,
      "grad_norm": 1.1837220191955566,
      "learning_rate": 5.016611295681063e-06,
      "loss": 0.2802,
      "step": 186
    },
    {
      "epoch": 2.7555555555555555,
      "grad_norm": 1.5560253858566284,
      "learning_rate": 4.983388704318937e-06,
      "loss": 0.3686,
      "step": 187
    },
    {
      "epoch": 2.7703703703703706,
      "grad_norm": 1.442020297050476,
      "learning_rate": 4.950166112956811e-06,
      "loss": 0.3648,
      "step": 188
    },
    {
      "epoch": 2.785185185185185,
      "grad_norm": 1.2866986989974976,
      "learning_rate": 4.916943521594685e-06,
      "loss": 0.335,
      "step": 189
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.9921827912330627,
      "learning_rate": 4.883720930232559e-06,
      "loss": 0.3052,
      "step": 190
    },
    {
      "epoch": 2.814814814814815,
      "grad_norm": 1.3171212673187256,
      "learning_rate": 4.850498338870432e-06,
      "loss": 0.3365,
      "step": 191
    },
    {
      "epoch": 2.8296296296296295,
      "grad_norm": 1.202288031578064,
      "learning_rate": 4.817275747508306e-06,
      "loss": 0.2726,
      "step": 192
    },
    {
      "epoch": 2.8444444444444446,
      "grad_norm": 1.5031251907348633,
      "learning_rate": 4.78405315614618e-06,
      "loss": 0.366,
      "step": 193
    },
    {
      "epoch": 2.859259259259259,
      "grad_norm": 1.2141317129135132,
      "learning_rate": 4.750830564784053e-06,
      "loss": 0.2861,
      "step": 194
    },
    {
      "epoch": 2.8740740740740742,
      "grad_norm": 1.64356529712677,
      "learning_rate": 4.717607973421927e-06,
      "loss": 0.3415,
      "step": 195
    },
    {
      "epoch": 2.888888888888889,
      "grad_norm": 1.2897413969039917,
      "learning_rate": 4.6843853820598015e-06,
      "loss": 0.3334,
      "step": 196
    },
    {
      "epoch": 2.9037037037037035,
      "grad_norm": 1.192264199256897,
      "learning_rate": 4.651162790697675e-06,
      "loss": 0.2927,
      "step": 197
    },
    {
      "epoch": 2.9185185185185185,
      "grad_norm": 1.36125648021698,
      "learning_rate": 4.617940199335549e-06,
      "loss": 0.3129,
      "step": 198
    },
    {
      "epoch": 2.9333333333333336,
      "grad_norm": 1.3408805131912231,
      "learning_rate": 4.5847176079734225e-06,
      "loss": 0.3277,
      "step": 199
    },
    {
      "epoch": 2.948148148148148,
      "grad_norm": 1.3519293069839478,
      "learning_rate": 4.551495016611296e-06,
      "loss": 0.2881,
      "step": 200
    },
    {
      "epoch": 2.962962962962963,
      "grad_norm": 1.2503290176391602,
      "learning_rate": 4.51827242524917e-06,
      "loss": 0.3282,
      "step": 201
    },
    {
      "epoch": 2.977777777777778,
      "grad_norm": 1.357706069946289,
      "learning_rate": 4.4850498338870435e-06,
      "loss": 0.3319,
      "step": 202
    },
    {
      "epoch": 2.9925925925925925,
      "grad_norm": 1.173913598060608,
      "learning_rate": 4.451827242524917e-06,
      "loss": 0.3099,
      "step": 203
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.173913598060608,
      "learning_rate": 4.451827242524917e-06,
      "loss": 0.2291,
      "step": 204
    },
    {
      "epoch": 3.0148148148148146,
      "grad_norm": 1.3551340103149414,
      "learning_rate": 4.418604651162791e-06,
      "loss": 0.3068,
      "step": 205
    },
    {
      "epoch": 3.0296296296296297,
      "grad_norm": 1.3206164836883545,
      "learning_rate": 4.3853820598006645e-06,
      "loss": 0.3416,
      "step": 206
    },
    {
      "epoch": 3.0444444444444443,
      "grad_norm": 1.417013168334961,
      "learning_rate": 4.352159468438539e-06,
      "loss": 0.2771,
      "step": 207
    },
    {
      "epoch": 3.0592592592592593,
      "grad_norm": 1.115646243095398,
      "learning_rate": 4.318936877076413e-06,
      "loss": 0.2828,
      "step": 208
    },
    {
      "epoch": 3.074074074074074,
      "grad_norm": 1.4733177423477173,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 0.3038,
      "step": 209
    },
    {
      "epoch": 3.088888888888889,
      "grad_norm": 1.3604916334152222,
      "learning_rate": 4.25249169435216e-06,
      "loss": 0.3128,
      "step": 210
    },
    {
      "epoch": 3.1037037037037036,
      "grad_norm": 1.6025428771972656,
      "learning_rate": 4.219269102990034e-06,
      "loss": 0.2808,
      "step": 211
    },
    {
      "epoch": 3.1185185185185187,
      "grad_norm": 1.5683448314666748,
      "learning_rate": 4.186046511627907e-06,
      "loss": 0.3538,
      "step": 212
    },
    {
      "epoch": 3.1333333333333333,
      "grad_norm": 1.5745620727539062,
      "learning_rate": 4.152823920265781e-06,
      "loss": 0.3071,
      "step": 213
    },
    {
      "epoch": 3.148148148148148,
      "grad_norm": 1.26665461063385,
      "learning_rate": 4.119601328903655e-06,
      "loss": 0.2568,
      "step": 214
    },
    {
      "epoch": 3.162962962962963,
      "grad_norm": 1.1559772491455078,
      "learning_rate": 4.086378737541528e-06,
      "loss": 0.2655,
      "step": 215
    },
    {
      "epoch": 3.1777777777777776,
      "grad_norm": 1.3667874336242676,
      "learning_rate": 4.053156146179402e-06,
      "loss": 0.2943,
      "step": 216
    },
    {
      "epoch": 3.1925925925925926,
      "grad_norm": 1.4076658487319946,
      "learning_rate": 4.0199335548172765e-06,
      "loss": 0.242,
      "step": 217
    },
    {
      "epoch": 3.2074074074074073,
      "grad_norm": 1.461408257484436,
      "learning_rate": 3.98671096345515e-06,
      "loss": 0.2792,
      "step": 218
    },
    {
      "epoch": 3.2222222222222223,
      "grad_norm": 1.3601031303405762,
      "learning_rate": 3.953488372093024e-06,
      "loss": 0.2788,
      "step": 219
    },
    {
      "epoch": 3.237037037037037,
      "grad_norm": 1.9735394716262817,
      "learning_rate": 3.9202657807308975e-06,
      "loss": 0.3367,
      "step": 220
    },
    {
      "epoch": 3.251851851851852,
      "grad_norm": 1.5413546562194824,
      "learning_rate": 3.887043189368771e-06,
      "loss": 0.2887,
      "step": 221
    },
    {
      "epoch": 3.2666666666666666,
      "grad_norm": 1.5698870420455933,
      "learning_rate": 3.853820598006645e-06,
      "loss": 0.3121,
      "step": 222
    },
    {
      "epoch": 3.2814814814814817,
      "grad_norm": 1.4351308345794678,
      "learning_rate": 3.8205980066445185e-06,
      "loss": 0.2859,
      "step": 223
    },
    {
      "epoch": 3.2962962962962963,
      "grad_norm": 1.5523098707199097,
      "learning_rate": 3.787375415282392e-06,
      "loss": 0.3496,
      "step": 224
    },
    {
      "epoch": 3.311111111111111,
      "grad_norm": 1.5026061534881592,
      "learning_rate": 3.754152823920266e-06,
      "loss": 0.2744,
      "step": 225
    },
    {
      "epoch": 3.325925925925926,
      "grad_norm": 1.565753698348999,
      "learning_rate": 3.72093023255814e-06,
      "loss": 0.2761,
      "step": 226
    },
    {
      "epoch": 3.3407407407407406,
      "grad_norm": 1.514609456062317,
      "learning_rate": 3.6877076411960135e-06,
      "loss": 0.2385,
      "step": 227
    },
    {
      "epoch": 3.3555555555555556,
      "grad_norm": 1.233588695526123,
      "learning_rate": 3.6544850498338876e-06,
      "loss": 0.2849,
      "step": 228
    },
    {
      "epoch": 3.3703703703703702,
      "grad_norm": 1.3447370529174805,
      "learning_rate": 3.621262458471761e-06,
      "loss": 0.3197,
      "step": 229
    },
    {
      "epoch": 3.3851851851851853,
      "grad_norm": 1.3895092010498047,
      "learning_rate": 3.588039867109635e-06,
      "loss": 0.2869,
      "step": 230
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.326200246810913,
      "learning_rate": 3.5548172757475086e-06,
      "loss": 0.2937,
      "step": 231
    },
    {
      "epoch": 3.414814814814815,
      "grad_norm": 1.2751388549804688,
      "learning_rate": 3.5215946843853823e-06,
      "loss": 0.2611,
      "step": 232
    },
    {
      "epoch": 3.4296296296296296,
      "grad_norm": 1.1317442655563354,
      "learning_rate": 3.4883720930232564e-06,
      "loss": 0.2733,
      "step": 233
    },
    {
      "epoch": 3.4444444444444446,
      "grad_norm": 1.3669071197509766,
      "learning_rate": 3.4551495016611296e-06,
      "loss": 0.2967,
      "step": 234
    },
    {
      "epoch": 3.4592592592592593,
      "grad_norm": 1.451213002204895,
      "learning_rate": 3.4219269102990037e-06,
      "loss": 0.3398,
      "step": 235
    },
    {
      "epoch": 3.474074074074074,
      "grad_norm": 2.680572748184204,
      "learning_rate": 3.3887043189368774e-06,
      "loss": 0.2973,
      "step": 236
    },
    {
      "epoch": 3.488888888888889,
      "grad_norm": 1.5603687763214111,
      "learning_rate": 3.355481727574751e-06,
      "loss": 0.2669,
      "step": 237
    },
    {
      "epoch": 3.5037037037037035,
      "grad_norm": 1.8472412824630737,
      "learning_rate": 3.322259136212625e-06,
      "loss": 0.2956,
      "step": 238
    },
    {
      "epoch": 3.5185185185185186,
      "grad_norm": 1.6677846908569336,
      "learning_rate": 3.2890365448504984e-06,
      "loss": 0.329,
      "step": 239
    },
    {
      "epoch": 3.533333333333333,
      "grad_norm": 1.8696357011795044,
      "learning_rate": 3.2558139534883724e-06,
      "loss": 0.2735,
      "step": 240
    },
    {
      "epoch": 3.5481481481481483,
      "grad_norm": 1.2237964868545532,
      "learning_rate": 3.222591362126246e-06,
      "loss": 0.3011,
      "step": 241
    },
    {
      "epoch": 3.562962962962963,
      "grad_norm": 1.2561728954315186,
      "learning_rate": 3.1893687707641198e-06,
      "loss": 0.2568,
      "step": 242
    },
    {
      "epoch": 3.5777777777777775,
      "grad_norm": 1.1599830389022827,
      "learning_rate": 3.156146179401994e-06,
      "loss": 0.3025,
      "step": 243
    },
    {
      "epoch": 3.5925925925925926,
      "grad_norm": 1.7549623250961304,
      "learning_rate": 3.122923588039867e-06,
      "loss": 0.3252,
      "step": 244
    },
    {
      "epoch": 3.6074074074074076,
      "grad_norm": 1.5652899742126465,
      "learning_rate": 3.089700996677741e-06,
      "loss": 0.341,
      "step": 245
    },
    {
      "epoch": 3.6222222222222222,
      "grad_norm": 1.7946059703826904,
      "learning_rate": 3.056478405315615e-06,
      "loss": 0.3244,
      "step": 246
    },
    {
      "epoch": 3.637037037037037,
      "grad_norm": 1.5564749240875244,
      "learning_rate": 3.0232558139534885e-06,
      "loss": 0.2686,
      "step": 247
    },
    {
      "epoch": 3.651851851851852,
      "grad_norm": 1.439971923828125,
      "learning_rate": 2.9900332225913626e-06,
      "loss": 0.3198,
      "step": 248
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 1.590828537940979,
      "learning_rate": 2.9568106312292363e-06,
      "loss": 0.2874,
      "step": 249
    },
    {
      "epoch": 3.6814814814814816,
      "grad_norm": 1.2883110046386719,
      "learning_rate": 2.9235880398671104e-06,
      "loss": 0.2475,
      "step": 250
    },
    {
      "epoch": 3.696296296296296,
      "grad_norm": 1.2013444900512695,
      "learning_rate": 2.8903654485049836e-06,
      "loss": 0.2911,
      "step": 251
    },
    {
      "epoch": 3.7111111111111112,
      "grad_norm": 1.5711296796798706,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 0.3113,
      "step": 252
    },
    {
      "epoch": 3.725925925925926,
      "grad_norm": 1.5199639797210693,
      "learning_rate": 2.8239202657807313e-06,
      "loss": 0.2462,
      "step": 253
    },
    {
      "epoch": 3.7407407407407405,
      "grad_norm": 1.2884533405303955,
      "learning_rate": 2.790697674418605e-06,
      "loss": 0.245,
      "step": 254
    },
    {
      "epoch": 3.7555555555555555,
      "grad_norm": 1.4477325677871704,
      "learning_rate": 2.7574750830564782e-06,
      "loss": 0.2409,
      "step": 255
    },
    {
      "epoch": 3.7703703703703706,
      "grad_norm": 1.2022032737731934,
      "learning_rate": 2.7242524916943523e-06,
      "loss": 0.2806,
      "step": 256
    },
    {
      "epoch": 3.785185185185185,
      "grad_norm": 1.467168927192688,
      "learning_rate": 2.691029900332226e-06,
      "loss": 0.2866,
      "step": 257
    },
    {
      "epoch": 3.8,
      "grad_norm": 2.0435290336608887,
      "learning_rate": 2.6578073089701e-06,
      "loss": 0.3211,
      "step": 258
    },
    {
      "epoch": 3.814814814814815,
      "grad_norm": 1.2129874229431152,
      "learning_rate": 2.6245847176079738e-06,
      "loss": 0.2765,
      "step": 259
    },
    {
      "epoch": 3.8296296296296295,
      "grad_norm": 1.4887040853500366,
      "learning_rate": 2.591362126245847e-06,
      "loss": 0.282,
      "step": 260
    },
    {
      "epoch": 3.8444444444444446,
      "grad_norm": 1.3326512575149536,
      "learning_rate": 2.558139534883721e-06,
      "loss": 0.307,
      "step": 261
    },
    {
      "epoch": 3.859259259259259,
      "grad_norm": 1.2139685153961182,
      "learning_rate": 2.5249169435215947e-06,
      "loss": 0.3053,
      "step": 262
    },
    {
      "epoch": 3.8740740740740742,
      "grad_norm": 1.6648317575454712,
      "learning_rate": 2.4916943521594684e-06,
      "loss": 0.2859,
      "step": 263
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 1.550366997718811,
      "learning_rate": 2.4584717607973425e-06,
      "loss": 0.3527,
      "step": 264
    },
    {
      "epoch": 3.9037037037037035,
      "grad_norm": 1.641121506690979,
      "learning_rate": 2.425249169435216e-06,
      "loss": 0.2735,
      "step": 265
    },
    {
      "epoch": 3.9185185185185185,
      "grad_norm": 1.244675874710083,
      "learning_rate": 2.39202657807309e-06,
      "loss": 0.3208,
      "step": 266
    },
    {
      "epoch": 3.9333333333333336,
      "grad_norm": 1.4783977270126343,
      "learning_rate": 2.3588039867109635e-06,
      "loss": 0.2869,
      "step": 267
    },
    {
      "epoch": 3.948148148148148,
      "grad_norm": 1.5731481313705444,
      "learning_rate": 2.3255813953488376e-06,
      "loss": 0.4189,
      "step": 268
    },
    {
      "epoch": 3.962962962962963,
      "grad_norm": 1.654645562171936,
      "learning_rate": 2.2923588039867112e-06,
      "loss": 0.2766,
      "step": 269
    },
    {
      "epoch": 3.977777777777778,
      "grad_norm": 1.6379601955413818,
      "learning_rate": 2.259136212624585e-06,
      "loss": 0.3227,
      "step": 270
    },
    {
      "epoch": 3.9925925925925925,
      "grad_norm": 1.3923341035842896,
      "learning_rate": 2.2259136212624586e-06,
      "loss": 0.3051,
      "step": 271
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.9158486127853394,
      "learning_rate": 2.1926910299003322e-06,
      "loss": 0.2157,
      "step": 272
    },
    {
      "epoch": 4.014814814814815,
      "grad_norm": 1.2792537212371826,
      "learning_rate": 2.1594684385382063e-06,
      "loss": 0.2568,
      "step": 273
    },
    {
      "epoch": 4.029629629629629,
      "grad_norm": 1.1235581636428833,
      "learning_rate": 2.12624584717608e-06,
      "loss": 0.2664,
      "step": 274
    },
    {
      "epoch": 4.044444444444444,
      "grad_norm": 1.2936333417892456,
      "learning_rate": 2.0930232558139536e-06,
      "loss": 0.2966,
      "step": 275
    },
    {
      "epoch": 4.059259259259259,
      "grad_norm": 1.2000664472579956,
      "learning_rate": 2.0598006644518273e-06,
      "loss": 0.2487,
      "step": 276
    },
    {
      "epoch": 4.074074074074074,
      "grad_norm": 1.4936013221740723,
      "learning_rate": 2.026578073089701e-06,
      "loss": 0.283,
      "step": 277
    },
    {
      "epoch": 4.088888888888889,
      "grad_norm": 1.551798939704895,
      "learning_rate": 1.993355481727575e-06,
      "loss": 0.275,
      "step": 278
    },
    {
      "epoch": 4.103703703703704,
      "grad_norm": 1.8705427646636963,
      "learning_rate": 1.9601328903654487e-06,
      "loss": 0.3045,
      "step": 279
    },
    {
      "epoch": 4.118518518518519,
      "grad_norm": 1.5146915912628174,
      "learning_rate": 1.9269102990033224e-06,
      "loss": 0.2575,
      "step": 280
    },
    {
      "epoch": 4.133333333333334,
      "grad_norm": 2.1336557865142822,
      "learning_rate": 1.893687707641196e-06,
      "loss": 0.293,
      "step": 281
    },
    {
      "epoch": 4.148148148148148,
      "grad_norm": 1.673853874206543,
      "learning_rate": 1.86046511627907e-06,
      "loss": 0.2228,
      "step": 282
    },
    {
      "epoch": 4.162962962962963,
      "grad_norm": 1.4035648107528687,
      "learning_rate": 1.8272425249169438e-06,
      "loss": 0.236,
      "step": 283
    },
    {
      "epoch": 4.177777777777778,
      "grad_norm": 1.5520694255828857,
      "learning_rate": 1.7940199335548175e-06,
      "loss": 0.2381,
      "step": 284
    },
    {
      "epoch": 4.192592592592592,
      "grad_norm": 1.8831232786178589,
      "learning_rate": 1.7607973421926911e-06,
      "loss": 0.3056,
      "step": 285
    },
    {
      "epoch": 4.207407407407407,
      "grad_norm": 1.524223804473877,
      "learning_rate": 1.7275747508305648e-06,
      "loss": 0.2288,
      "step": 286
    },
    {
      "epoch": 4.222222222222222,
      "grad_norm": 1.7227911949157715,
      "learning_rate": 1.6943521594684387e-06,
      "loss": 0.3001,
      "step": 287
    },
    {
      "epoch": 4.237037037037037,
      "grad_norm": 1.6464650630950928,
      "learning_rate": 1.6611295681063126e-06,
      "loss": 0.2767,
      "step": 288
    },
    {
      "epoch": 4.2518518518518515,
      "grad_norm": 1.462727665901184,
      "learning_rate": 1.6279069767441862e-06,
      "loss": 0.2583,
      "step": 289
    },
    {
      "epoch": 4.266666666666667,
      "grad_norm": 1.3373020887374878,
      "learning_rate": 1.5946843853820599e-06,
      "loss": 0.1861,
      "step": 290
    },
    {
      "epoch": 4.281481481481482,
      "grad_norm": 1.3069071769714355,
      "learning_rate": 1.5614617940199335e-06,
      "loss": 0.2279,
      "step": 291
    },
    {
      "epoch": 4.296296296296296,
      "grad_norm": 2.0804059505462646,
      "learning_rate": 1.5282392026578074e-06,
      "loss": 0.3373,
      "step": 292
    },
    {
      "epoch": 4.311111111111111,
      "grad_norm": 1.5276566743850708,
      "learning_rate": 1.4950166112956813e-06,
      "loss": 0.2211,
      "step": 293
    },
    {
      "epoch": 4.325925925925926,
      "grad_norm": 1.7238777875900269,
      "learning_rate": 1.4617940199335552e-06,
      "loss": 0.2492,
      "step": 294
    },
    {
      "epoch": 4.340740740740741,
      "grad_norm": 1.7892625331878662,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 0.2521,
      "step": 295
    },
    {
      "epoch": 4.355555555555555,
      "grad_norm": 1.83062744140625,
      "learning_rate": 1.3953488372093025e-06,
      "loss": 0.2445,
      "step": 296
    },
    {
      "epoch": 4.37037037037037,
      "grad_norm": 1.665940284729004,
      "learning_rate": 1.3621262458471762e-06,
      "loss": 0.2188,
      "step": 297
    },
    {
      "epoch": 4.385185185185185,
      "grad_norm": 1.7071560621261597,
      "learning_rate": 1.32890365448505e-06,
      "loss": 0.2294,
      "step": 298
    },
    {
      "epoch": 4.4,
      "grad_norm": 1.5902904272079468,
      "learning_rate": 1.2956810631229235e-06,
      "loss": 0.2443,
      "step": 299
    },
    {
      "epoch": 4.4148148148148145,
      "grad_norm": 1.6071012020111084,
      "learning_rate": 1.2624584717607974e-06,
      "loss": 0.2268,
      "step": 300
    },
    {
      "epoch": 4.42962962962963,
      "grad_norm": 1.481217384338379,
      "learning_rate": 1.2292358803986712e-06,
      "loss": 0.2245,
      "step": 301
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 1.7338011264801025,
      "learning_rate": 1.196013289036545e-06,
      "loss": 0.2481,
      "step": 302
    },
    {
      "epoch": 4.459259259259259,
      "grad_norm": 1.83079195022583,
      "learning_rate": 1.1627906976744188e-06,
      "loss": 0.2771,
      "step": 303
    },
    {
      "epoch": 4.474074074074074,
      "grad_norm": 1.6256608963012695,
      "learning_rate": 1.1295681063122925e-06,
      "loss": 0.2418,
      "step": 304
    },
    {
      "epoch": 4.488888888888889,
      "grad_norm": 1.7368595600128174,
      "learning_rate": 1.0963455149501661e-06,
      "loss": 0.2325,
      "step": 305
    },
    {
      "epoch": 4.503703703703704,
      "grad_norm": 1.8986256122589111,
      "learning_rate": 1.06312292358804e-06,
      "loss": 0.2685,
      "step": 306
    },
    {
      "epoch": 4.518518518518518,
      "grad_norm": 1.9530152082443237,
      "learning_rate": 1.0299003322259137e-06,
      "loss": 0.2734,
      "step": 307
    },
    {
      "epoch": 4.533333333333333,
      "grad_norm": 2.2950448989868164,
      "learning_rate": 9.966777408637875e-07,
      "loss": 0.2775,
      "step": 308
    },
    {
      "epoch": 4.548148148148148,
      "grad_norm": 1.7885551452636719,
      "learning_rate": 9.634551495016612e-07,
      "loss": 0.2299,
      "step": 309
    },
    {
      "epoch": 4.562962962962963,
      "grad_norm": 1.8106021881103516,
      "learning_rate": 9.30232558139535e-07,
      "loss": 0.2665,
      "step": 310
    },
    {
      "epoch": 4.5777777777777775,
      "grad_norm": 1.7523531913757324,
      "learning_rate": 8.970099667774087e-07,
      "loss": 0.2514,
      "step": 311
    },
    {
      "epoch": 4.592592592592593,
      "grad_norm": 1.5606251955032349,
      "learning_rate": 8.637873754152824e-07,
      "loss": 0.216,
      "step": 312
    },
    {
      "epoch": 4.607407407407408,
      "grad_norm": 1.3957760334014893,
      "learning_rate": 8.305647840531563e-07,
      "loss": 0.2133,
      "step": 313
    },
    {
      "epoch": 4.622222222222222,
      "grad_norm": 1.4215636253356934,
      "learning_rate": 7.973421926910299e-07,
      "loss": 0.2159,
      "step": 314
    },
    {
      "epoch": 4.637037037037037,
      "grad_norm": 1.9139257669448853,
      "learning_rate": 7.641196013289037e-07,
      "loss": 0.2799,
      "step": 315
    },
    {
      "epoch": 4.651851851851852,
      "grad_norm": 1.5956664085388184,
      "learning_rate": 7.308970099667776e-07,
      "loss": 0.2386,
      "step": 316
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 1.7055957317352295,
      "learning_rate": 6.976744186046513e-07,
      "loss": 0.2256,
      "step": 317
    },
    {
      "epoch": 4.681481481481481,
      "grad_norm": 2.2495598793029785,
      "learning_rate": 6.64451827242525e-07,
      "loss": 0.2795,
      "step": 318
    },
    {
      "epoch": 4.696296296296296,
      "grad_norm": 1.4281206130981445,
      "learning_rate": 6.312292358803987e-07,
      "loss": 0.2295,
      "step": 319
    },
    {
      "epoch": 4.711111111111111,
      "grad_norm": 1.7233028411865234,
      "learning_rate": 5.980066445182725e-07,
      "loss": 0.2586,
      "step": 320
    },
    {
      "epoch": 4.725925925925926,
      "grad_norm": 2.274329662322998,
      "learning_rate": 5.647840531561462e-07,
      "loss": 0.2477,
      "step": 321
    },
    {
      "epoch": 4.7407407407407405,
      "grad_norm": 1.6101205348968506,
      "learning_rate": 5.3156146179402e-07,
      "loss": 0.2304,
      "step": 322
    },
    {
      "epoch": 4.7555555555555555,
      "grad_norm": 1.9844639301300049,
      "learning_rate": 4.983388704318938e-07,
      "loss": 0.2888,
      "step": 323
    },
    {
      "epoch": 4.770370370370371,
      "grad_norm": 1.7177938222885132,
      "learning_rate": 4.651162790697675e-07,
      "loss": 0.2543,
      "step": 324
    },
    {
      "epoch": 4.785185185185185,
      "grad_norm": 1.7451181411743164,
      "learning_rate": 4.318936877076412e-07,
      "loss": 0.2362,
      "step": 325
    },
    {
      "epoch": 4.8,
      "grad_norm": 1.9255115985870361,
      "learning_rate": 3.9867109634551497e-07,
      "loss": 0.2689,
      "step": 326
    },
    {
      "epoch": 4.814814814814815,
      "grad_norm": 1.6680419445037842,
      "learning_rate": 3.654485049833888e-07,
      "loss": 0.2505,
      "step": 327
    },
    {
      "epoch": 4.82962962962963,
      "grad_norm": 1.8660964965820312,
      "learning_rate": 3.322259136212625e-07,
      "loss": 0.2671,
      "step": 328
    },
    {
      "epoch": 4.844444444444444,
      "grad_norm": 1.9916861057281494,
      "learning_rate": 2.9900332225913623e-07,
      "loss": 0.2519,
      "step": 329
    },
    {
      "epoch": 4.859259259259259,
      "grad_norm": 1.5888378620147705,
      "learning_rate": 2.6578073089701e-07,
      "loss": 0.2505,
      "step": 330
    },
    {
      "epoch": 4.874074074074074,
      "grad_norm": 1.654725193977356,
      "learning_rate": 2.3255813953488374e-07,
      "loss": 0.2466,
      "step": 331
    },
    {
      "epoch": 4.888888888888889,
      "grad_norm": 1.4593290090560913,
      "learning_rate": 1.9933554817275749e-07,
      "loss": 0.2039,
      "step": 332
    },
    {
      "epoch": 4.9037037037037035,
      "grad_norm": 1.5956478118896484,
      "learning_rate": 1.6611295681063126e-07,
      "loss": 0.2221,
      "step": 333
    },
    {
      "epoch": 4.9185185185185185,
      "grad_norm": 1.33909273147583,
      "learning_rate": 1.32890365448505e-07,
      "loss": 0.192,
      "step": 334
    },
    {
      "epoch": 4.933333333333334,
      "grad_norm": 1.7868711948394775,
      "learning_rate": 9.966777408637874e-08,
      "loss": 0.2477,
      "step": 335
    }
  ],
  "logging_steps": 1,
  "max_steps": 335,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 150,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4351170969600.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}